The AI Action Plan Meets Local Government Reality

The views expressed are those of the author and do not necessarily reflect the views of ASPA as an organization.

By Mohammad M. Rahman
June 5, 2026

As governments face staffing shortages, rising service demands and aging administrative systems, artificial intelligence is rapidly moving from experimentation to operational use in public service delivery. Federal agencies are expanding AI applications to improve administrative efficiency, while state and local governments are pursuing digital modernization to enhance service delivery. At the same time, concerns about fairness, transparency and accountability are growing as automated systems become more embedded in decisions that directly affect citizens.

This tension reflects a central governance challenge for public administrators. AI offers opportunities to improve performance and reduce administrative burden, but it also raises difficult questions about equity, oversight and democratic accountability. Federal guidance on AI governance, including guidance from the Office of Management and Budget, has emphasized risk management, transparency and human oversight in high-impact decision systems.

For public administrators, these issues are no longer theoretical. Agencies are increasingly expected to modernize services while protecting fairness and the rights of vulnerable populations.

The Efficiency Case for AI in Government

Public agencies operate under persistent workforce shortages, budget constraints and rising expectations for faster and more accessible services. Staff processing unemployment insurance, SNAP benefits and housing assistance often handle large volumes of repetitive documentation, eligibility verification and citizen inquiries. These demands contribute to processing delays, administrative backlogs and employee burnout.

AI tools can help address these operational challenges. Automated systems can organize records, detect incomplete applications and route cases more efficiently. AI-enabled chat systems can respond to routine questions outside business hours, improving access for residents who might otherwise face long wait times or complex phone systems. Internal tools can assist staff by summarizing case files and reducing repetitive administrative tasks. These applications can improve operational efficiency and allow human staff to focus on cases requiring discretion, judgment or direct human interaction.

For citizens, the stakes are significant. Delays in public benefits often translate into immediate hardship for families relying on food assistance, housing support or unemployment insurance. When implemented carefully, AI has the potential to reduce administrative friction and improve responsiveness in public programs.

Equity, Bias and Accountability Risks

Unlike the private sector, government is responsible not only for efficiency but also for fairness, transparency and equal treatment under the law. This distinction becomes critical when automated systems influence eligibility decisions or enforcement actions.

AI systems are trained on historical data, and public-sector datasets often reflect long-standing inequities, inconsistent enforcement or administrative bias. Without safeguards, these patterns can be reproduced or amplified in automated decision-making systems. The central concern is not only technical error but also accountability, specifically whether citizens can understand how decisions are made and who is responsible when systems fail.

A widely cited example comes from Michigan’s unemployment insurance system, where an automated fraud detection tool incorrectly flagged thousands of residents, leading to wrongful penalties and financial harm. Investigations documented in reporting such as ProPublica’s analysis of Michigan’s system highlight how weak oversight in automated systems can produce large-scale harm.

These risks are especially pronounced for individuals with limited digital literacy, language barriers or limited internet access. Even well-designed systems can produce unequal outcomes if agencies do not continuously evaluate impacts across demographic groups. Broader concerns about algorithmic fairness are also evident in tools such as COMPAS, a risk assessment system widely studied for potential disparities in predictive outcomes across racial groups, as documented in ProPublica’s investigation of COMPAS. These cases continue to shape national debates about transparency, fairness and accountability in algorithmic governance.

Local Government Capacity and the Accountability Gap

Local governments face a distinct challenge in governing AI systems due to limited technical capacity and resource constraints. Many municipalities rely heavily on external vendors for software solutions while lacking internal expertise to audit, validate or explain algorithmic systems.

This creates an accountability gap. When automated systems influence public decisions, citizens still hold government responsible regardless of whether the technology was developed internally or procured externally. If agencies cannot clearly explain how systems function or how decisions are reviewed, public trust may erode.

The Government Accountability Office has noted that many public agencies lack sufficient documentation, testing standards and oversight capacity for emerging AI systems. These gaps are even more pronounced at the local level, where staffing and funding constraints are more severe. Without stronger governance capacity, agencies risk delegating essential public functions to systems they cannot fully evaluate or control.

Governance Principles for Responsible AI

Artificial intelligence should be treated as a governance issue, not simply a technology issue. Public administrators must balance innovation with accountability to ensure that modernization strengthens rather than weakens public trust.

Several principles are central to responsible AI governance in the public sector. Human oversight should remain central in high-impact decisions involving benefits, eligibility or enforcement. Automated systems should support administrative decision-making, not replace it.

Transparency is essential. Governments should clearly communicate when AI systems are used in public programs and explain their role in decision-making processes in plain language.

Equity monitoring should be continuous. Agencies must evaluate whether automated systems produce disparate impacts across demographic or socioeconomic groups and adjust systems accordingly.

Vendor accountability must also be strengthened. Procurement contracts should require documentation, auditability and performance standards that allow agencies to evaluate system behavior over time.

These safeguards are not barriers to innovation. They are necessary conditions for ensuring that innovation aligns with democratic governance.

Conclusion: The Future of AI in Public Administration

The central question for public administrators is no longer whether AI will be used in government, but how it will be governed. Artificial intelligence will continue to reshape public service delivery, but its legitimacy depends on the strength of oversight, transparency and accountability frameworks.

Efficiency gains alone are not sufficient. The core challenge is whether governments can modernize service delivery while maintaining fairness, accountability and public trust. As agencies continue adopting AI systems, the decisions made today will shape the credibility and effectiveness of public administration in the decades ahead.

---

Author: Mohammad M. Rahman is pursuing a Ph.D. in Strategic Leadership Studies at the School of Strategic Leadership Studies in the College of Business at James Madison University. His interdisciplinary research interests include public administration, higher education policy, nonprofit management and leadership, and organizational psychology. He also serves as an adjunct faculty member in the College of Business at James Madison University.

 

(10 votes, average: 4.70 out of 5)
Loading...